Nanoscale Redox Reaction at Metal/Oxide Interface: A Case Study on Schottky Contact and ReRAM [1 ed.] 9784431548492, 9784431548508

Citation preview

NIMS Monographs

Takahiro Nagata

Nanoscale Redox Reaction at Metal/Oxide Interface A Case Study on Schottky Contact and ReRAM

NIMS Monographs Series Editor Naoki OHASHI, National Institute for Materials Science, Tsukuba, Ibaraki, Japan Editorial Board Takahito OHMURA, National Institute for Materials Science, Tsukuba, Ibaraki, Japan Yoshitaka TATEYAMA, National Institute for Materials Science, Tsukuba, Ibaraki, Japan Takashi TANIGUCHI, National Institute for Materials Science, Tsukuba, Ibaraki, Japan Kazuya TERABE, National Institute for Materials Science, Tsukuba, Ibaraki, Japan Masanobu NAITO, National Institute for Materials Science, Tsukuba, Ibaraki, Japan Nobutaka HANAGATA, National Institute for Materials Science, Tsukuba, Japan Kenjiro MIYANO, National Institute for Materials Science, Tsukuba, Ibaraki, Japan

NIMS publishes specialized books in English covering from principle, theory and all recent application examples as NIMS Monographs series. NIMS places a unity of one study theme as a specialized book which was specialized in each particular field, and we try for publishing them as a series with the characteristic (production, application) of NIMS. Authors of the series are limited to NIMS researchers. Our world is made up of various “substances” and in these “materials” the basis of our everyday lives can be found. Materials fall into two major categories such as organic/polymeric materials and inorganic materials, the latter in turn being divided into metals and ceramics. From the Stone Ages - by way of the Industrial Revolution - up to today, the advance in materials has contributed to the development of humankind and now it is being focused upon as offering a solution for global problems. NIMS specializes in carrying out research concerning these materials. NIMS: http://www.nims.go.jp/eng/index.html

More information about this series at http://www.springer.com/series/11599

Takahiro Nagata

Nanoscale Redox Reaction at Metal/Oxide Interface A Case Study on Schottky Contact and ReRAM

123

Takahiro Nagata Research Center for Functional Materials National Institute for Materials Science Tsukuba, Ibaraki, Japan

ISSN 2197-8891 ISSN 2197-9502 (electronic) NIMS Monographs ISBN 978-4-431-54849-2 ISBN 978-4-431-54850-8 (eBook) https://doi.org/10.1007/978-4-431-54850-8 © National Institute for Materials Science, Japan 2020 This work is subject to copyright. All rights are reserved by the National Institute for Materials Science, Japan (NIMS), whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of applicable copyright laws and applicable treaties, and permission for use must always be obtained from NIMS. Violations are liable to prosecution under the respective copyright laws and treaties. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. NIMS and the publisher make no warranty, express or implied, with respect to the material contained herein. This Springer imprint is published by the registered company Springer Japan KK part of Springer Nature. The registered company address is: Shiroyama Trust Tower, 4-3-1 Toranomon, Minato-ku, Tokyo 105-6005, Japan

Preface

Today, integrated circuit technologies are at a crucial turning point in the establishment of the fundamentals for further advancement. To overcome the performance limits of conventional materials such as the SiO2 gate, polycrystalline Si gate, and Al wires, it is necessary to develop a new material with a new functionality not yet found in Si devices. Oxide materials are a good candidate to replace Si devices since these materials show exotic properties in accordance with the composition design and/or doping technique. These materials should realize future functional devices with high-k, ferroelectric, magnetic, and optical properties. In this book, we discuss the investigation and intentional control of the metal/oxide interface structure and electrical properties using the data obtained by nondestructive methods such as X-ray photoelectron spectroscopy (XPS). At the metal/oxide interface, oxygen plays an important role in redox reactions, and thus affects the electrical properties. To observe and control these properties, a metal Schottky contact for an optical sensor application (Chaps. 1 and 2) and a metal/oxide resistive random-access memory structure (Chaps. 3 and 4) are investigated. For example, nitrogen plasma treatment on an oxide surface can reduce the surface electron accumulation layer, resulting in an enhancement of the Schottky property of a metal/oxide interface. Additionally, in Chap. 5, we briefly introduce combinatorial thin-film synthesis, which is our specialty and a suitable method for exploring new materials. While we do not attempt to cover every single aspect of oxide research in this book, we do aim to present discussions on selected topics that are both representative and possibly of technological interest. We expect this book to be of interest to scientists and engineers working in the field of metal oxides. Tsukuba, Japan

Takahiro Nagata

v

Acknowledgments

First of all, I would like to acknowledge the editorial teams of Springer and NIMS Monographs for their patience and valuable suggestions. I am deeply grateful to my many colleagues at NIMS, the BL15XU NIMS Beamline, and Meiji University for invaluable suggestions, discussions, and technical support. I am also grateful to HiSOR, Hiroshima University, and JAEA/SPring-8 for developing HX-PES at BL15XU, SPring-8. The HX-PES measurements were performed under the approval of the NIMS Beamline Station (BL-15XU) (Proposal nos. 2007B4604, 2009A4600, 2009B4601, 2010A4604, 2010B4600, 2011A4611, 2011B4611, and 2012A4613). Part of the work in this book was supported by Kakenhi Grant-in-Aid for Scientific Research B 19760224, a Grant-in-Aid for Key Technology, “Atomic Switch Programmed Device”, and the World Premier International Research Center Initiative (WPI), from Japan’s Ministry of Education, Culture, Sports, Science, and Technology (MEXT) Japan.

vii

Contents

1 General Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 3

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy Contacts on Single-Crystal ZnO . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Interface Formation and Characterization . . . . . . . . . . . . . . . 2.2.1 Schottky Barrier Height . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Gibbs Free Energy: Ellingham Diagram . . . . . . . . . . 2.2.3 Phase Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.4 Characterization: Electrical Measurements . . . . . . . . . 2.2.5 Characterization: X-Ray Photoelectron Spectroscopy . 2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts on Polar Face of ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Crystal Structural Analysis . . . . . . . . . . . . . . . . . . . . 2.3.2 Surface Morphology . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Electrical Properties . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Chemical Bonding States (HX-PES Measurements) . . 2.3.5 Surface Termination Effect . . . . . . . . . . . . . . . . . . . . 2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

5 5 6 6 7 9 10 11

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

15 15 19 19 20 23 24 24

3 Surface Passivation Effect on Schottky Contact Formation of Oxide Semiconductors . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Near-Atmospheric-Pressure Nitrogen Plasma Treatment . 3.2.1 Atmospheric-Pressure Nitrogen Plasma Source . . 3.2.2 Nitridation of Oxide Surface . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

27 27 27 28 30

. . . . .

. . . . .

. . . . .

ix

x

Contents

3.3 Near-Atmospheric-Pressure Nitrogen Plasma Passivation . 3.3.1 Nitridation of ZnO Surface . . . . . . . . . . . . . . . . . 3.3.2 Improvement of Metal/ZnO Interface . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

32 32 34 38

. . . . . . . .

. . . . . . . .

41 41 42 42 44 46 46 49

..

55

..

56

..

58

.. ..

58 59

4 Bias-Induced Interfacial Redox Reaction in Oxide-Based Resistive Random-Access Memory Structure . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Nanoionic-Type ReRAM Structure . . . . . . . . . . . . . . . . . . . . . 4.2.1 Sample Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2 Electrical Properties of Cu/HfO2/Pt Structure . . . . . . . . . 4.3 HX-PES Measurements Under Bias Application . . . . . . . . . . . . 4.3.1 Cu/HfO2 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.2 Pt/HfO2 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Filament Formation Process in Cu/HfO2/Pt and Pt/HfO2/Pt ReRAM Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5 Bias-Induced Cu Migration Behavior in Cu/HfO2 ReRAM Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6 Effect of Bottom Electrode on Interfacial Structure and Switching Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6.1 Electrical Properties of Pt/Cu/HfO2/Pt/Si and Pt/Cu/HfO2/TiN/Si Structures . . . . . . . . . . . . . . . . . 4.6.2 Interfacial Structure Between Cu and HfO2 . . . . . . . . . . 4.6.3 Correlation Between Ion Migration and Switching Degradation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6.4 Effect of Bottom Electrode on Conductive Filament Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

..

61

.. .. ..

62 63 64

5 Switching Control of Oxide-Based Resistive Random-Access Memory by Valence State Control of Oxide . . . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Valence Control Scheme . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Combinatorial Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 Ta–Nb Binary Oxide System . . . . . . . . . . . . . . . . 5.3.2 Valence State of Oxides . . . . . . . . . . . . . . . . . . . . 5.3.3 Electrical Properties . . . . . . . . . . . . . . . . . . . . . . . 5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

69 69 69 70 70 71 72 74 74

6 Combinatorial Thin-Film Synthesis for New Nanoelectronics Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Combinatorial Thin-Film Synthesis . . . . . . . . . . . . . . . . . . . . . . .

75 75 75

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

Contents

6.3 Focused Ar Ion-Beam Sputtering for Combinatorial Synthesis . 6.3.1 Energy of Focused Ar Ion Beam Sputtering . . . . . . . . 6.3.2 Metal Thin-Film Growth on Oxide . . . . . . . . . . . . . . . 6.3.3 Combinatorial Thin-Film Synthesis by FIBS . . . . . . . . 6.4 Combinatorial Characterization . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Two-Dimensional X-Ray Diffraction Method . . . . . . . 6.4.2 Atomic Force Microscopy-Based Electrical Property Mapping Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

. . . . . .

79 79 81 82 82 82

... ... ...

84 85 86

7 General Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

89

. . . . . .

. . . . . .

Chapter 1

General Introduction

The integrated circuit technologies stand at a crucial turning point in establishing the fundamentals for further advancement. To overcome the performance limits of conventional materials such as SiO2 gate, polycrystalline Si gate, and Al wiring, it is necessary to develop a new material with a new functionality that is not found in Si devices up to now, such as nonvolatile memory function. Various material combinations and device structures have been proposed and investigated regardless of in-organic or organic materials. In this field, our interest is in in-organic compound materials, especially oxides. From the point of view of the electrical device application, oxide ceramic and crystals are widely used in bulk form for more than half a century, such as ceramics capacitor, electro-optical switch, surface acoustic wave devices, and so on. In the film case, the practical application in electrical devices was limited in the use of a transparent conducting oxide (TCO) due to its transparency in the visible range and relatively high electrical conductivity except for SiO2 [1, 2]. In the late 1990s, hafnium oxide-based high-k oxide and InGaZnOx (IGZO) accelerated the thin-film oxide application in the electrical devices [3, 4]. High-k oxide replaces the SiO2 gate material and enhances the scale down rule limitation, with the so-called Moore technology. In contrast, IGZO archived the lower power consumption in the liquid crystal display application due to its transparency and mobility instead of the Si transistor. These applications focused spotlight on other oxide materials. In the case of semiconductor, wideband gap oxides such as SnO2 , ZnO, and In2 O3 exhibit a highly sensitive surface which has proven useful as a sensor technology [5–7]. Recently, additional various electrical applications such as electrode materials in displays [8], light-emitting diodes (LEDs) [9], and transparent thin-film transistors (TTFTs) [10, 11] have also been investigated. From the point of view of the multifunctional oxide materials, recently resistive random-access memory (ReRAM) has been proposed as a new application for oxide materials due to its simple structure. An oxide sandwiched between two metal electrodes shows reversible electric field-induced resistance switching behaviors. For nonvolatile memory applications, ferroelectric materials have been researched for two decades. However, typical ferroelectric materials indicated the scale effect, and some of the ferroelectric materials, including the hazardous or alkali metals such © National Institute for Materials Science, Japan 2020 T. Nagata, Nanoscale Redox Reaction at Metal/Oxide Interface, NIMS Monographs, https://doi.org/10.1007/978-4-431-54850-8_1

1

2

1 General Introduction

as Pb, Li, and Ca, are not suitable for the Si-based film electronics. In contrast, oxide-based ReRAM structure indicated material flexibility and scale down ability. Several types of ReRAM structure have been demonstrated. A typical resistive switching model is based on a thermal effect initiated by a voltage-induced partial dielectric breakdown that forms a discharge filament modified by Joule heating [12, 13]. The intrinsic material properties also induce changes in resistance. For example, the insulator–metal transition in perovskite oxides such as (Pr,Ca)MnO3 [14–16] and SrTiO3 :Cr [17] is induced by electronic charge injection operations such as doping. These materials indicate the scale down ability that contributes to the high-density device integration. For these thin-film nanoelectronics device applications, with decreasing device scale, the importance of interface structure is pronounced, which is strongly related to the formation of the oxygen vacancies. At the oxide semiconductor surface, oxygen vacancies induced the Fermi-level pinning, which can be controlled by post deposition treatments such as plasma treatment. At the metal/oxide interface, oxygen plays an important role in redox reactions, affecting the electrical properties. To observe and control them, metal Schottky contact for an optical sensor application and metal/oxide resistive random-access memory structure are investigated. In this book, the investigation and intentional control of metal/oxide interface structure and electrical properties with the data obtained by nondestructive methods such as Xray photoelectron spectroscopy (XPS) are discussed; for example, nitrogen plasma treatment on an oxide surface. These treatments can reduce the surface electron accumulation layer, resulting in an enhancement of Schottky property of a metal/oxide interface. This book consists of five chapters based on our research results as follows: Chapters 2 and 3: The interface structure of Schottky metal on oxide semiconductor. As an example, we chose zinc oxide (ZnO) as an oxide semiconductor since ZnO is a polar material. The metal and oxygen terminated surface can be obtained, which is suitable for the investigation of the oxygen effects on the interface. Chapters 4 and 5: The oxygen effect on the switching properties of ReRAM structure. As an example, we chose hafnium oxide (HfO2 )-based ReRAM structure. HfO2 is used as a high-k gate insulator for advanced complementary metal/oxide semiconductor (CMOS) technologies; it has shown resistance switching phenomena, and there has been of increased interest in the use of HfO2 and related oxides as potential ReRAM materials. Chapter 6: Brief introduction of combinatorial thin-film synthesis In the former chapters, the combinatorial thin-film synthesis plays an important role for systematic and high-throughput analysis. To help readers understand the sample fabrication and analysis, the combinatorial thin-film synthesis is introduced briefly. While the book does not attempt to cover every single aspect of oxides research, it does aim to present discussions on selected topics that are both representative and possibly of technological interest.

References

3

References 1. Minami T (2005) Transparent conducting oxide semiconductors for transparent electrodes. Semicond Sci Technol 20:S35. https://doi.org/10.1088/0268-1242/20/4/004 2. Ginley DS, Bright C (2000) Transparent conducting oxides. MRS Bull 25:15. https://doi.org/ 10.1557/mrs2000.256 3. Hosono H, Kikuchi N, Ueda N, Kawazoe H (1996) Working hypothesis to explore novel wide band gap electrically conducting amorphous oxides and examples. J Non-Cryst Solids 198–200:165. https://doi.org/10.1016/0022-3093(96)80019-6 4. Nomura K, Ohta H, Ueda K, Kamiya T, Hirano M, Hosono H (2003) Thin-film transistor fabricated in single-crystalline transparent oxide semiconductor. Science 300:1269. https:// doi.org/10.1126/science.1083212 5. Kohnke EE (1962) Electrical and optical properties of natural stannic oxide crystals. J Phys Chem Solids 23:1557. https://doi.org/10.1016/0022-3697(62)90236-6 6. Nagasawa M, Shionoya S, Makishim S (1965) Vapor reaction growth of SnO2 single crystals and their properties. Jpn J Appl Phys 4:195. https://doi.org/10.1143/JJAP.4.195 7. Choudhary J, Ogale SB, Shinde SR, Kulkarni VN, Vendatesan T, Harshavardhan KS, Strikovski M, Hannoyer B (2004) Pulsed-electron-beam deposition of transparent conducting SnO2 films and study of their properties. Appl Phys Lett 84:1483. https://doi.org/10.1063/1.1651326 8. Batzill M, Katsiev K, Burst JM, Diebold U, Chaka AM, Delley B (2005) Gas-phase-dependent properties of SnO2 (110), (100), and (101) single-crystal surfaces: structure, composition, and electronic properties. Phys Rev B 72:165414. https://doi.org/10.1103/PhysRevB.72.165414 9. Anisimov OV, Gaman VI, Maksimova NK, Mazalov SM, Chernikov EV (2006) Electrical and gas-sensitive properties of a resistive thin-film sensor based on tin oxide. Semiconductors 40:704. https://doi.org/10.1134/S1063782606060170 10. Kim H, Pique A, Horwitz JS, Mattoussi H, Murata H, Kafafi ZH, Chrisey DB (1999) Indium tin oxide thin films for organic light-emitting devices. Appl Phys Lett 74:3444. https://doi.org/ 10.1063/1.124122 11. von Wenckstern H, Splith D, Lanzinger S, Schmidt F, Müller S, Schlupp P, Karsthof R, Grundmann M (2015) pn-hetero diodes with n-type In2 O3 . Adv Electr Mater 1:1400026. https://doi. org/10.1002/aelm.201400026 12. Pagnia H, Sotnik N (1988) Bistable switching in electroformed metal–insulator–metal devices. Phys Stat Sol (a) 108:11. https://doi.org/10.1002/pssa.2211080102 13. Chudnovskii FA, Odynets LL, Pergament AL, Stefanovich GB (1996) Electroforming and switching in oxides of transition metals: the role of metal–insulator transition in the switching mechanism. J Solid State Chem 122:95. https://doi.org/10.1006/jssc.1996.0087 14. Asamitsu A, Tomioka Y, Kuwahara H, Tokura Y (1997) Current switching of resistive states in magnetoresistive manganites. Nature 388:50. https://doi.org/10.1038/40363 15. Fors R, Khartsev SI, Grishin AM (2005) Giant resistance switching in metal-insulatormanganite junctions: evidence for Mott transition. Phys Rev B 71:045305. https://doi.org/ 10.1103/PhysRevB.71.045305 16. Kim DS, Kim YH, Lee CE, Kim YT (2006) Colossal electroresistance mechanism in a Au/Pr0.7 Ca0.3 MnO3 /Pt sandwich structure: evidence for a Mott transition. Phys Rev B 74:174430. https://doi.org/10.1103/physrevb.74.174430 17. Meijer GI, Staub U, Janousch M, Johnson SL, Delley B, Neisius T (2005) Valence states of Cr and the insulator-to-metal transition in Cr-doped SrTiO3 . Phys Rev B 72:155102. https://doi. org/10.1103/PhysRevB.72.155102

Chapter 2

Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy Contacts on Single-Crystal ZnO

2.1 Introduction Zinc oxide (ZnO), a wide band-gap II–VI semiconductor, has major potential for use in optical devices. For example, it has attracted much attention for potential use in light-emitting and light-detecting devices in the ultraviolet (UV) region [1–3]. Although similar devices based on gallium nitride (GaN) have been put to practical use, ZnO has a number of advantages over GaN, including higher quantum efficiency, high exciton binding energy at room temperature, greater resistance to high-energy radiation, and amenability to wet chemical etching [4]. High-quality and thermally reliable Schottky contacts are crucial. We have proposed a UV-region Schottky-type photodiode using ZnO [5]. Compared with the formation of an ohmic contact on ZnO, the formation of a Schottky contact on ZnO is complicated, owing to the very high donor concentration in the surface region, which consists of native defects such as oxygen vacancies and zinc interstitials [6]. Allen et al. prepared a high-quality AgO Schottky contact on ZnO by using a conventional simple surface cleaned by organic solvents [7]. Au and Pt Schottky contacts on bulk ZnO wafers have shown good Schottky properties [8], but the electrical properties of metal contacts on ZnO changed with time or after annealing [9, 10]. The effects of the interface oxidization on electrical properties have been discussed [11]. Both pretreatments and metal/ZnO interfaces are essential for realizing a high-quality Schottky contact. Furthermore, the Schottky barrier height (SBH) is related to the work function of materials. High-density interface defects reduce the metal work function [12], suggesting that an understanding of the metal/ZnO interfaces is the key to controlling the properties of Schottky contacts. ZnO has a wurtzite crystal structure with two distinct {0001} planes, as shown in Fig. 2.1. A lack of inversion symmetry and ionic bonds make this material polar. The Zn-terminated plane (0001) and the O-terminated plane (000–1) are, respectively, denoted as the Zn-polar face and the O-polar face. These two faces have different structures, compositions, and chemical and physical properties [13, 14]. The interface structure and thermal stabilities of metal/ZnO substrates should therefore be different. © National Institute for Materials Science, Japan 2020 T. Nagata, Nanoscale Redox Reaction at Metal/Oxide Interface, NIMS Monographs, https://doi.org/10.1007/978-4-431-54850-8_2

5

6

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.1 Schematic illustration of the crystal structure of ZnO

Here, we report on the SBH behavior of Pt–Ru alloy contacts on n-ZnO substrates at both the Zn-polar and O-polar faces. Pt and Ru are, respectively, used as high and low work function materials. Furthermore, the Gibbs free energy of Pt oxide is larger than that of Ru oxide, indicating that the Ru oxide layer is more easily formed. Using the combination of polarity, work function, and oxidization energy differences, we investigated the interface structure effects on the SBH behavior by the combinatorial synthesis technique.

2.2 Interface Formation and Characterization 2.2.1 Schottky Barrier Height Figure 2.2 shows the energy-level diagram for a metal and n-type oxide semiconductor before and after contact formation. At the metal/semiconductor interface, in the absence of defect states, Schottky barrier height (Φ B ) is given by Φ B = ΦM − ΦS ,

(2.1)

where Φ M is the metal work function and Φ S is the semiconductor work function. Φ B is also given by Φ S = χ + (E C − E F ),

(2.2)

2.2 Interface Formation and Characterization

7

Fig. 2.2 Schematic energy-level diagrams for a an ideal metal/oxide interface prior to contacting, and b metal contact on an oxide film. E FM : Fermi-level of the metal, E F : Fermi-level of the oxide, E C : conduction band, E V : valence band, Eg: band-gap, Φ m : work function of the metal, X: electron affinity of the oxide

with the electron affinity χ , the Fermi-level E F , and the conduction band edge E C . For an n-type semiconductor, the Fermi-level is close to the conduction band, and for Φ S ≈ χ , we approximate Φ B = Φ M −χ .

(2.3)

However, an actual metal/oxide interface shows a smaller Φ B than that calculated using Eq. (2.1). The metal/oxide interface is affected by various phenomena, such as oxidization and defect formation.

2.2.2 Gibbs Free Energy: Ellingham Diagram To use oxides and metal/oxide heterostructures, the Gibbs free energy (ΔG) is useful. Gibbs free energy is a thermodynamic potential that measures the thermodynamic

8

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

driving force that makes a reaction (oxidization) to occur. ΔG is given by G = H − T S

(2.4)

where ΔH is the enthalpy, T is the absolute temperature, and ΔS is the entropy. ΔH is the actual energy. If it is negative, the reaction is exoergic. In contrast, if it is positive, the reaction is endoergic. ΔS is a measure of the disorder of a system. In the case of oxide formation from a metal, a negative value for ΔG means that oxidization can proceed spontaneously without external inputs. The standard Gibbs free energy of formation of a compound (oxide) is the change in Gibbs free energy that accompanies the formation of 1 mol of a substance in its standard state from its constituent elements in their standard states (the most stable form of the element at 1 bar of pressure and the specified temperature, usually 298.15 K or 25 °C). An Ellingham diagram plots the standard free energy of a reaction as a function of temperature, as shown in Fig. 2.3 [15]. Since ΔH and ΔS are essentially constant with temperature unless a phase change occurs, the plot of free energy versus temperature can be drawn as a series of straight lines, where ΔS is the slope and ΔH is the y-intercept. The slope of the line changes when any of the materials involved melt or vaporize. The free energy of formation is negative

Fig. 2.3 Ellingham diagram

2.2 Interface Formation and Characterization

9

for most metal oxides, and so the diagram is drawn with ΔG = 0 at the top, and the values of ΔG shown are all negative numbers. The Ellingham diagram shown is for metals reacting to form oxides. The oxygen partial pressure is taken as 1 atmosphere, and all the reactions are normalized to consume 1 mol of O2 . By using the diagram, the standard free energy change for any included reaction can be found at any temperature. Along with enabling the calculation of the equilibrium composition of the system, the data on the diagram is useful in other ways, as we shall see.

2.2.3 Phase Diagram The SBH is related to the work function of the metal. It is generally known that the work function of a metal is affected by the surface potential, which means an effect of the orientation of the metal layer, such as tungsten [16–18]. In the case of tungsten, the values of the work functions of (111), (100), and (110) faces are 4.47, 4.63, and 5.25 eV, respectively. Additionally, the work function of a metal is affected by interface defects. A high density of defects reduces the value of the work function [19]. To control the Schottky barrier height, the crystal structures of the Schottky metal layer should be investigated. For understanding and controlling the crystal structure of alloys, a phase diagram is beneficial. In the case of Pt–W as shown in Fig. 2.4, there is no intermediate stoichiometric alloy, meaning that the Pt–W alloy has a possibility of linear controllability of the work function. Actually, the combinatorial synthesis of the Pt–W alloy revealed the continuously changing behavior of the work function, as shown in Fig. 2.5. The Pt–W alloy has 111-oriented structures. The work function decreases continuously with increasing W content.

Fig. 2.4 Binary phase diagram of Pt–W alloy

10

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.5 a Sample picture, b XRD mapping, and c relative work functions measured by KFM of Pt–W alloy composition spread sample

2.2.4 Characterization: Electrical Measurements A typical way to investigate Φ B is by electrical measurements such as current–voltage (I–V) and capacitance–voltage (C–V) measurements. From I–V measurements, both Φ B and the ideality factor (n) were determined using Eqs. (2.5) and (2.6), and are consistent with thermionic emission theory in excess of several kT/q [20].  I = I0 exp

 q(V − I Rs ) −1 , nkT

(2.5)

2.2 Interface Formation and Characterization

I0 = A A∗∗ T 2 exp

11



 −qΦ B , kT

(2.6)

where I 0 is the saturation current, Rs is the series resistance, k is Boltzmann’s constant, T is the absolute temperature, A is the contact area, q is the electronic charge, and A** is the effective Richardson constant. From C–V measurements, the majority carrier distribution n(x) is given by [21, 22] n(x) = −

  C 3 dC −1 , qkε0 dV

(2.7)

where kε0 is the semiconductor permittivity, q is the electron charge, and x = kε0 A/C is the test junction space charge layer width at the applied biasing voltage V. Using the majority carrier distribution, Φ B can be estimated.

2.2.5 Characterization: X-Ray Photoelectron Spectroscopy X-ray photoelectron spectroscopy (XPS) is an electron spectroscopy method. X-rays eject electrons from inner-shell orbitals of the elements that exist within a material; then, an energy analyzer detects the photoelectrons. The kinetic energy (E K ) of the photoelectrons is determined by the energy of the X-ray (hν) and the electron binding energy (E B ) as E K = hν − E B .

(2.8)

The experimentally measured binding energy is given by E B = hν − E K − ϕspec ,

(2.9)

where ϕ spec is the work function of the spectrometer. The electron binding energies are dependent on the chemical environment of the atom, making XPS useful for identifying the chemical bonding state (oxidation state) and electronic state of materials. The X-ray penetration depth is at least several micrometers. However, the electrons leave with very low energy, so they can only escape from the top few nanometers. Tanuma, Powell, and Penn developed the Tanuma–Powell–Penn (TPP-2 M) equation to calculate the inelastic mean free paths (IMFPs: λ) using Green’s function [23]. The TPP-2 M equation indicated that IMFP can become longer by using high-energy X-rays. In our study, a combination of conventional X-ray photoelectron spectroscopy using Al Kα radiation (Al-XPS: hν = 1486.6 eV) and hard X-ray photoelectron spectroscopy (HX-PES: hν = 5.95 keV) was used to investigate the

12

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.6 Probability of photoelectron detection of SnO2 based on IMFP calculation at various takeoff angles (TOAs). The solid and dashed lines show conventional XPS and HX-PES, respectively. The inset shows the measurement setup

interface structure. For example, the IMFPs for the Sn 3d5/2 core-level spectra of Al-XPS and HX-PES for SnO2 calculated using the TPP-2 M equation are 1.8 and 6.9 nm, respectively, indicating that HX-PES probes three times deeper than Al-XPS. Furthermore, to determine the details of band bending behavior, angle-resolved XPS (AR-XPS) measurements were carried out. While crossing the material, the photoelectrons are subject to the laws of absorption. Thus, the probability (I 0 ) that the emitted photoelectron reaches the surface can be estimated against a given depth (x) of photoemission. Figure 2.6 shows an example of the probability (I0 = exp(−x/λ)) for the Sn 3d5/2 core-level spectra at various measurement angles as a function of depth. This plot indicates that, by changing the measurement angle, the probing depth can be controlled and it is approximately 3 × λ [24, 25]. In our experiments, HX-PES was performed at the SPring-8 BL15XU undulator beamline with a 200 mm mean radius spectrometer (VG Scienta R4000) [26]. The total energy resolutions for the Al-XPS and HX-PES measurements were estimated to be ~600 and ~230 meV, respectively. To determine the absolute binding energy, the photoelectron spectroscopy data was calibrated against the Fermi-level position of Au. The sample was in contact with the system ground, whose energy was equal to the Fermi-level position of Au, via a conductive copper tape. To analyze the HX-PES results, peaks were fitted using the Voigt function with the Doniach-Šunji´c function after the background had been removed by employing the Shirley function [27, 28]. At the site of metal electrode formation, XPS can describe the band offset and bending of a metal/oxide interface. Since the core levels have a fixed binding energy difference from the conduction and valence band edges, they can be used to trace shifts of the band edges with respect to the Fermi level, as shown in Fig. 2.7. Furthermore, HX-PES probes at 20 nm below the surface, indicating the bulk property of ZnO. Figure 2.8 shows the Zn 2p3/2 spectra of the Zn-polar ZnO substrate and 10 nm Pt/ZnO interface obtained by HX-PES. The Pt electrode formation moved

2.2 Interface Formation and Characterization

13

Fig. 2.7 Correlation of valence band and core spectra in photoelectron spectroscopy and energy band diagram of Pt/ZnO interface

Zn 2p3/2 to a lower binding energy, meaning that the Fermi-level moved to the middle of the band-gap of ZnO at the interface. Furthermore, the take-off angle (TOA) dependence of the Zn 2p3/2 position revealed the upward band bending of the ZnO at the interface, shown in Fig. 2.7, suggesting the formation of a depletion region at the ZnO surface. These results are consistent with Schottky contact formation at the metal/semiconductor interface.

14

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.8 a Zn 2p3/2 core spectra of a Zn-polar ZnO substrate and a Pt/ZnO interface at a TOA of 88°. b TOA dependence of the binding energy of the Zn 2p3/2 core spectra. The squares show the Pt/ZnO interface. The Zn 2p3/2 core spectrum of a Zn-polar ZnO substrate at a TOA of 88°, which probes a similar depth to the others in terms of the IMFP, is also indicated as a dashed line

For measurements of metal work function, photoelectron spectroscopy is also useful. Under even atmospheric pressure conditions, small numbers of low-energy photoelectrons can be detected. This photoelectron yield spectroscopy (PYS) can estimate not only the yield spectrum but also DOS as UPS and XPS, in air ambient. Figure 2.9 shows a photoelectron yield spectrum of the Ru film on a glass substrate. The ionization energy is defined as the energy difference between the vacuum level and the extrapolated edge of the Fermi-level of the metal. Therefore, the work function can be estimated by linearly extrapolating the edge of the square-root photoelectron yield to the baseline. The work function of the Ru film is 4.85 ± 0.1 eV, which is comparable to previously reported values (4.6–4.7 eV) [16–18].

2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts …

15

Fig. 2.9 Photoelectron yield spectrum of a Ru film on a glass substrate

2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts on Polar Face of ZnO Figure 2.10a shows a schematic of the sample structure fabricated by combinatorial synthesis technique. The details of combinatorial synthesis technique are shown in Chap. 6. An ohmic contact layer was made by the Al lift-off method. Circular top Schottky contacts of 130 μm diameter were deposited by the composition spread method. Pt and Ru were deposited as Schottky metals. Figure 2.10b shows the PSY measurement of a Pt–Ru alloy on a glass substrate. The work functions of the Pt–Ru composition spread sample change continuously.

2.3.1 Crystal Structural Analysis The 2D XRD images of the Pt and Ru films showed spotted peaks at approximately 40° (Fig. 2.11a). The peaks were assigned to a cubic structure with (111) reflection (Pt) and a hexagonal structure with (0002) reflection (Ru). The peaks of the electrodes are observed at approximately 40°. These peaks shift to higher angles with increasing Ru content and have two possible crystal structures. One is the Pt phase with a cubic structure, and the other structure is the Ru phase with a hexagonal structure. The electrodes with a Pt-type structure show only the (111) reflection, indicating a 111-oriented film. Those with a Ru-type structure show only the (0002) reflection, indicating a 0001-oriented film. The Pt–Ru alloy phase diagram has two intermediate phase lines of crystal structures changing at Pt content of approximately 20 and 39 at.%, as shown in Fig. 2.11b [29]. At a Pt content below 20 at.%, the crystal structure of the Pt–Ru alloy is hexagonal; at ≥39 at.%, it is cubic; between 20 and 39 at.%, it is a mixed structure phase. The rocking curves of the Pt

16

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.10 a Schematic illustration of the Pt–Ru composition spread sample. b Work functions measured by PYS of the Pt–Ru composition spread film on a glass substrate

phase with cubic structure (111) reflections or the Ru phase with hexagonal structure (0002) reflections are plotted as a function of Pt content, as shown in Fig. 2.11c. FWHM increased with decreasing Pt content, and showed a peak at a Pt content of 29 at.%. This change is suitable for the Pt–Ru alloy phase diagram, because the Pt content of 29 at.% results in a mixed structure phase. Mixed structure phases may become a low crystalline phase. From the results of the ω-2θ XRD and ψ scans, it cannot be determined whether the film is of the Pt phase or the Ru phase. X-ray pole figure analysis was also performed. The X-ray pole figure was measured at 3° intervals of scan step in φ angle. The ranges of the 2θ and ψ angles are detected simultaneously from 30 to 60° and from 25 to 75°, respectively. Figure 2.12a–c shows the results of Pt content −

of 0, 29, and 100 at.%, respectively. The value of ψ was 42.7° for the (10 1 2) plane of ZnO. In Fig. 2.12a, peaks with a six-fold symmetry at 60° in the φ scan can −

be observed at ψ = 61° for the (10 1 1) plane of Ru. This implies that Ru with a hexagonal structure grew on the ZnO substrate with its c-axis normal to the substrate surface epitaxially. Figure 2.12c shows peaks with a six-fold symmetry at ψ = 54.7°

2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts …

17

Fig. 2.11 a 2D-XRD images of the Pt and Ru films on the O-polar face. b Dependence of peak positions and χ-FWHMs on Pt content for Pt (111) reflection with a cubic structure or Ru (0002) reflection with a hexagonal structure on Zn-polar (solid squares) and O-polar (open squares) faces. c Pt–Ru alloy phase diagram

18

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.12 X-ray pole figures of a Ru, b Pt content of 29 at.%, c Pt electrodes on ZnO, and d theoretical pole figures of Ru and Pt on ZnO

for the (200) plane of Pt, although the peaks of the 111-oriented epitaxial Pt film showed a normal three-fold symmetry. This result suggests that the Pt phase with a rotation of 60° was also grown on ZnO epitaxially. At the point of 29 at.% on the mixed structure phase, the X-ray pole figure shows a mixed pattern of Ru and Pt phases. These results reveal that the crystal structures of the Pt–Ru alloy film on ZnO change from the Pt phase to the Ru phase at a Pt content of approximately 29 at.%, as in the case of the Pt–Ru alloy phase diagram, and that the Pt–Ru alloy on a ZnO was grown on the ZnO substrate epitaxially.

2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts …

19

Fig. 2.13 AFM images of a Pt and b Ru films on Zn-polar face, and c Pt and d Ru films on O-polar face

2.3.2 Surface Morphology The Pt–Ru alloy film surfaces were observed by AFM, as shown in Fig. 2.13. Both polar faces showed similar surface changes. The Pt films had small-grained structures with a root mean square (RMS) value of 1.72 nm at the Zn-polar face. In contrast, the low-Pt films showed flat surfaces with an RMS value of 0.18 nm for Ru at the Zn-polar face. The maximum RMS difference between Zn-polar and O-polar faces was 0.07 nm. Thus, the surface roughness difference should have a negligible effect on the surface potential.

2.3.3 Electrical Properties The I–V measurements revealed that the Schottky properties become dominant with increasing Pt content of the films. Below 13 at.%, in the Ru structure phase, ohmic behaviors become dominant. To investigate SBH, we modeled the I–V characteristics

20

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.14 Pt content dependences of SBH (Φ B : squares) and ideality factor (n: circles) of the Pt– Ru alloy films on Zn-polar (solid marks) and O-polar (open marks) faces calculated from the I–V characteristic. The inset shows the I–V characteristic of Pt–Ru alloy contact on the O-polar face

of all Pt–Ru alloy contacts using Eqs. (2.4) and (2.5). In this study, the theoretical value of A** = 32 A cm−2 K−2 was used [23]. Figure 2.14 shows the dependence of SBH and the ideality factor on Pt content. At both polar faces, SBH decreased with decreasing Pt content, suggesting that the metal work function changed continuously. SBH changed by 0.11 eV at the Zn-polar face and by 0.09 eV at the O-polar face. SBH was greater at the Zn-polar face than at the O-polar face except at a Pt content of 44 at.%. The ideality factors on the O-polar face are larger than that on the Zn-polar face. Furthermore, with increasing Ru content, the ideality factor increases. High ideality factor means an increase of the recombination current instead of the ideal diffusion current, suggesting the defect formation at the interface.

2.3.4 Chemical Bonding States (HX-PES Measurements) To investigate the electronic states of the Pt–Ru/ZnO interface, we performed HXPES measurements at different compositions of 10 nm thick Pt–Ru alloy films on the Zn-polar and O-polar faces. Figure 2.15a shows the Pt 4f and Ru 3d core-level HX-PES spectra for Pt and Ru at the O-polar face at a TOA of 88° from the surface.

2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts …

21

Fig. 2.15 a Typical Pt 4f and Ru 3d core-level HX-PES spectra of Pt and Ru on the O-polar face at a TOA of 88°. Solid lines, measured spectra; open circles, sum-fitted curves. Dashed lines are fitted curves for each bond: Pt, PtO, Ru, and RuO2 . b Dependence of the HX-PES spectrum intensity ratio of oxide to metal on Pt and Ru content (circles, PtO/Pt; triangles, RuO2 /Ru) at a TOA of 88°. Solid symbols, Pt–Ru alloy films on Zn-polar faces; open symbols, on O-polar faces

In the Pt film, the peaks at the binding energies of 72.3 and 75.6 eV correspond to metallic Pt, and small peaks at 73.5 and 76.8 eV are attributed to PtO [30, 31]. In the Ru film, peaks at 281.3 and 285.5 eV correspond to metallic Ru, and small peaks at 282.5 and 286.7 eV are attributed to RuO2 [32, 33]. Figure 2.15b shows the relative intensities of the metals and the oxides as a function of the Pt or Ru content. The intensity ratio of the films was higher on the O-polar face than on the Zn-polar face. These results suggest that the Pt–Ru alloy films on the O-polar face were more oxidized. With decreasing Pt content, the PtO/Pt ratio increases, and the RuO2 /Ru ratio decreases slightly, indicating that the alloy film, which is in a mixed phase and poorly crystalline, has the most oxidized layer in the Pt–Ru alloy films. At a TOA of 20°, a clear difference in the Zn 2p3/2 spectrum was confirmed, as shown in Fig. 2.16. Intensities of the Zn 2p3/2 spectra were normalized to the corresponding Pt or Ru core-level peak intensities. The Zn 2p3/2 intensity of the Zn-polar face is much greater than that of the O-polar face, suggesting that the Zn

22

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

Fig. 2.16 (Color online) Zn 2p3/2 spectra of Pt and Ru films on O-polar and Zn-polar faces at a TOA of 20°. Solid lines, measured spectra; open circles, sum-fitted curves. Dashed lines are fitted curves for each bond: ZnO, ZnOx , ZnPt, ZnPtO, ZnPtOx , ZnRuO, and ZnRuOx

in the Zn-polar face is present at a shallower location than in the O-polar face. To confirm this consideration, javascript:goWordLink(%22maintenance%22) the IMFP of photoelectrons was calculated using the TPP-2 M equation. The Pt/ZnO interface structure was modeled using two different structures. One was PtO for the Pt/O-polar ZnO interface, and the other was PtZn for the Pt/Zn-polar interface. The IMFPs of PtZn and PtO are 5.26 and 6.11 nm, respectively. The IMFP of PtO is longer than that of PtZn. If the top of Zn layers of both polar faces were present at a similar depth within the difference of one monolayer length, the intensity of the Zn 2p for PtO would appear greater than that for PtZn. This means that the Zn in the Zn-polar face is present at a shallower location than in the O-polar face, indicating that Zn diffused into the Pt layer. Moreover, unlike the O-polar face, three additional peaks were identified in the Zn-polar face, as shown in Fig. 2.16b: ZnPt, ZnPtO, and ZnPtOx. These species were assigned considering the difference in electronegativity of each atom and in reference to previous studies [34, 35]. Some groups reported that the Zn 2p peak of the ZnO bond of zinc ferrite and zinc titanate, in which the coordination number of Zn is six, shifted to a higher binding energy than the Zn 2p peak of four coordinate Zn atoms in oxides. In our case, ZnPt and ZnPt alloys below a Pt content of 48% have a cubic structure in which the coordination number of Zn is six [36]. Furthermore, some of the ZnPt area is likely to be oxidized. The Zn in oxidized ZnPt, such as Zn2 PtO4, also has a coordination number of six. Thus, the peaks at higher binding energy were assigned as ZnPtO and ZnPtOx bonds. For the Ru film, the

2.3 Combinatorial Synthesis of Binary Alloy Metal Contacts …

23

component at the lowest binding energy corresponding to a metal–Zn bond was not identified. The Ru film was more oxidized than the Pt film.

2.3.5 Surface Termination Effect The metal work functions of Pt and Ru are 5.32–5.50 and 4.6–4.7 eV, respectively, and the electron affinity of ZnO is 4.1–4.4 eV [37–39]. Thus, SBH between the Pt–Ru alloy and ZnO should be in the range of 1.1–0.3 eV. Furthermore, the Znpolar face appears to show a slightly higher SBH than the calculated value, owing to polarization; this bends the band edge upward. The O-polar face should show the opposite property. In the I–V measurement, the SBH increased with increasing Pt content, and the Znpolar face showed a slightly higher SBH than the O-polar face. However, the SBHs were smaller than the calculated values, and the films with the Ru phase showed ohmic properties. According to the crystal structural analysis and HX-PES results, these phenomena can be explained by the complex interaction of three factors: (A) oxidization, (B) Zn- or O-termination, and (C) crystal structure of the films on the SBH. (A) Oxidization: In the Ru phase, the alloys on both polar faces of ZnO showed ohmic properties, and the O-polar face showed poor electrical properties. These phenomena were caused by the oxidization of the metal layer. The HX-PES results indicate that the Pt–Ru film was more oxidized on the O-polar face than on the Zn-polar face. In addition, the films were in a mixed phase with poor crystallinity, suggesting that they had the thickest oxidized layer of the Pt–Ru alloy films. Shiraishi et al. reported that oxygen at the interface lowers the metal’s Fermilevel, and defects in the interface induced by the interface oxidization decrease the metal’s work function [40]. This theoretical result agrees closely with our result. (B) Zn- or O-termination: The HX-PES results revealed the relationship between the interface structures and the chemical states of the films on the Zn- and Opolar faces. At the Zn-polar face, Zn diffused into the metal layer, which implies a difference in oxidization at the interface. The Zn diffusion was expected to prevent the oxidization of metal, since Zn is used as a deoxidant in alloys. The Zn-polar face appears to inhibit the oxidization of the contact more than the Opolar face. However, the work function of Pt (5.32–5.50 eV) is likely to decrease owing to the metal work function of Zn (3.74 eV). Therefore, SBH decreases. At the O-polar face, in contrast, metal oxide species form thickly around the interface. Thus, the SBH of the O-polar face is smaller than that of the Zn-polar face. (C) Crystal structure: The changes in SBH of Pt–Ru alloy electrodes on the Zn-polar and O-polar faces were 0.11 and 0.09 eV, respectively. The difference in SBH between the highest and lowest values is smaller than the value calculated

24

2 Changes in Schottky Barrier Height Behavior of Pt–Ru Alloy …

from the work function. If the change in SBH is continuous, the difference in SBH from the Pt film to the Pt–Ru film with a Pt content of 44 at.% should be 0.44 eV. However, SBH decreased slightly (Fig. 2.15) and changed the ohmic contact at the Ru phase. Some groups reported that the metal work function changing behavior of alloys shows some dependence on the phase of the alloy [41, 42]. The XRD and ellipsometric measurements indicate that the crystal structure affects the electrical properties: the Pt–Ru alloy films with Schottky behaviors showed the cubic structure (Pt phase) and the same optical properties at a Pt content of >39 at.%.

2.4 Summary Pt–Ru alloy composition spread films were deposited on ZnO single-crystal substrates as Schottky contacts by combinatorial ion-beam deposition. The crystal structures of the Pt–Ru alloy changed from cubic to hexagonal phase with increasing Pt content, as seen in the bulk Pt–Ru alloy phase diagram. The SBH of the films in the Pt phase decreased with decreasing Pt content irrespective of the polar face of the ZnO substrate, although the Pt–Ru alloy with a Pt content of